Data-Driven Planning for Short-Haul Heavy-Duty Electric Vehicles: Insights from Naturalistic Driving Patterns

Abstract

Severe emissions from the transportation sector in the United States has encouraged growth in the use of electric vehicles. The adoption of heavy-duty battery electric vehicles (HDBEVs), however, has been delayed by uncertainties in their range and operational capabilities. These factors, coupled with a high initial cost of ownership, have presented a significant barrier to fleet operators considering electrifying their fleet. To address these concerns, having access to a variety of empirical data is crucial for fleet operators to make informed decisions about the viability of HDBEVs. Naturalistic driving data for diesel Class 8 trucks, which is collected on the basis that drivers should ideally behave as they would under current driving conditions, can be leveraged to model the behavior of HDBEVs under comparable conditions. This thesis investigates the use of telemetry data from two naturalistic driving studies for Class 8 vehicles, along with data for the current electric vehicle charging infrastructure, to assess the viability of using HDBEVs to fulfill routes for fleet operations. It also presents a tool that (1) can be used by individuals to evaluate which HDBEVs would be viable replacements for diesel vehicles on a workday shift basis and (2) where charging infrastructure would be necessary to support them. Previous efforts to plan for HDBEV charging infrastructure improvements in the United States have largely ignored the existing charging infrastructure, which can provide a solid foundation for implementing HDBEV charging solutions. This work provides a framework for evaluating the current charging infrastructure, predicting the performance of HDBEVs with real-world driving data, and identifying gaps in the charging network that should be addressed for the efficient and balanced implementation of HDBEVs in fleet applications.